T2 Mapping: how many echoes does it require?

Abstract

Purpose: The purpose of this study is to determine the difference between T2 relaxation times obtained from varying numbers of echoes. The results from phantoms and cartilage are compared between the varying number of echoes. Methods: Phantoms comprised of varying concentrations of copper sulfate were placed on the anterior of the knee, just distal to the patella and offset laterally. The knee and phantom were then imagedd using a dual echo sequence, as well as multiple-echoe sequences. All imaging was performed on a GE Signa HDx 1.5T. Cartilage on one slice of the series was then identified, as well as the phantoms. T2 maps were calculated from the dual-echo using a least-squares linear fit to the natural log of the signal values. For the multiple-echo sequence, two methods were used - one where all the echoes were used in the estimate of T2, and one where the first echo was dropped. For very short T2 relaxation times (less than 2× the first echo time), all echoes were used for both cases. The mean and standard deviation of the T2 relaxation times in each of the regions was then calculated. Results: Results are shown in the table below. The image shows an example of the T2 relaxation time for the cartilage calculated using the multiple-echo sequence.Tabled 1Calculated T2 Relaxation TimesDual-EchoDropped First EchoAll EchoesMeanStdDevMeanStdDevMeanStdDevPhantom 131.213.017.06.116.93.8Phantom 241.94.844.29.339.14.0Phantom 391.47.991.121.577.38.1Phantom 4240.375.7159.950.2131.020.4Cartilage50.319.437.517.436.213.9 Open table in a new tab Conclusions: Stimulated echoes play a large part in the variability of T2 measurements obtained from multiple-echo sequences. The effect of stimulated echoes can be mitigated in the pulse sequence by applying crusher gradients (GE’s Cartigram sequence), or by ignoring the first echo in sequences that do not attempt to reduce the stimulated echoes. The disadvantage of ignoring the first echo is that it contains the most signal, and tissue with relatively short T2 relaxation times will not be estimated accurately. Using a dual-echo sequence, with a second echo that is not a multiple of the first echo, also eliminates the effect of stimulated echoes, but provides an estimate of T2 which is much more susceptible to noise. The values estimated for T2 are over-estimated for lower values of T2 because the second echo will hit the noise floor. The values are also over-estimated for areas of lower signal intensity for the same reason. This explains why the phantom values are very close between the multiple-echo sequence and the dual-echo sequence in the range of 40–90 ms, but the cartilage values in the same range are over-estimated by the dual-echo sequence.